YaST1 + YaST2 = YaST3?
I'm sure this has been discussed to some extent here, nonetheless, I would like to get an idea of where SuSE are going with YaST. There are some features which I really liked in YaST which are missing in YaST2. The biggest thing I find lacking is the ability to determine what is currently installed on my system while browsing the updates. For example, if I go to the KDE update directory on the SuSE ftp site, I would like to know which applications are already on my system. We used to be able to do this with YaST. Another thing I find a bit disorienting is the way the contents of the /etc/rc.config have been moved into /etc/sysconfig. This certainly has its good points. It is easier to find a particular configuration item this way. I suspect it also lays the groundwork for bigger and better things to come. One of these is, I hope, a return to providing a mask over these configuration items. Now it seems necessary to start processes such as Apache by explicitly modifying the runlevels. That's fine, if that's what SuSE intend for us to do. The interface we have for that is likewise very nice. The difficult part seems to be figuring out what SuSE expect us to do. That is, *should* I be using the runlevel editor for starting services automatically? Is there a paradigm shift in YaST? Am I just having trouble adjusting to a new and better way of doing things, or are there really shortcomings in the current (SuSE 8). I very much like the new look and feel of the GUI based YaST2. The terminal based version, well....(how 'bout using tab to move the focus?) What do others think of YaST verses YaST2? Please note, this is NOT an exercise in SuSE bashing. In a world where the integrity of corporations seem to be vanishing faster than the forests, I can honestly say my trust in SuSE is rock solid. This is the only operating system I chose to use. Any other that I use is due to circumstances of employment. Steven
On Fri, Jul 05, Steven T. Hatton wrote:
Another thing I find a bit disorienting is the way the contents of the /etc/rc.config have been moved into /etc/sysconfig. This certainly has its good points. It is easier to find a particular configuration item this way. I suspect it also lays the groundwork for bigger and better things to come. One of these is, I hope, a return to providing a mask over these configuration items.
What are you missing? You can edit this variables in three ways: 1. vi/emacs/your favorite editor 2. with the YaST2 config module for the service to which the sysconfig file belongs. 3. YaST2 sysconfig editor (like YaST2 rc.config editor before). You have the same three choices as before.
Now it seems necessary to start processes such as Apache by explicitly modifying the runlevels. That's fine, if that's what SuSE intend for us to do. The interface we have for that is likewise very nice. The difficult part seems to be figuring out what SuSE expect us to do. That is, *should* I be using the runlevel editor for starting services automatically?
You can do what you wish. If you like to do it with the mouse, use the YaST2 runlevel editor. If you don't like this, use insserv or chkconfig on the console. Or use ln and rm, ...
Is there a paradigm shift in YaST? Am I just having trouble adjusting to a new and better way of doing things, or are there really shortcomings in the current (SuSE 8).
We only make some things easier: why creating/deleting links and edit START_* variables, if creating/deleting links is enough and more flexible? Now you can start services only for one runlevel and not, as before, for all or none.
I very much like the new look and feel of the GUI based YaST2. The terminal based version, well....(how 'bout using tab to move the focus?) What do others think of YaST verses YaST2?
I always use tab to move to the next focus. Thorsten -- Thorsten Kukuk http://www.suse.de/~kukuk/ kukuk@suse.de SuSE Linux AG Deutschherrnstr. 15-19 D-90429 Nuernberg -------------------------------------------------------------------- Key fingerprint = A368 676B 5E1B 3E46 CFCE 2D97 F8FD 4E23 56C6 FB4B
On Friday 05 July 2002 13:15, Thorsten Kukuk wrote:
On Fri, Jul 05, Steven T. Hatton wrote:
[snip]
I very much like the new look and feel of the GUI based YaST2. The terminal based version, well....(how 'bout using tab to move the focus?) What do others think of YaST verses YaST2?
I always use tab to move to the next focus.
Even in the "Runlevel editor: details" window, where it takes 16 tabs to cycle through the whole window? I realy mis a shortcut there that takes me to the services list.
Thorsten
Leen
On Friday 05 July 2002 04:15, Thorsten Kukuk wrote:
On Fri, Jul 05, Steven T. Hatton wrote:
Another thing I find a bit disorienting is the way the contents of the /etc/rc.config have been moved into /etc/sysconfig. This certainly has its good points. It is easier to find a particular configuration item this way. I suspect it also lays the groundwork for bigger and better things to come. One of these is, I hope, a return to providing a mask over these configuration items.
What are you missing? You can edit this variables in three ways: 1. vi/emacs/your favorite editor
That would be XEmacs. I have done just that. Part of the problem with doing things of this nature is that I'm never sure exactly what SuSE's configuration facilities would have me do. If there is a 'SuSE' way to make a change, I want to use it, rather than go in and modify a file by hand. Don't get me wrong. I like to know what SuSEconfig is doing to the config files, and I'm familiar with many of the config files.
2. with the YaST2 config module for the service to which the sysconfig file belongs.
Perhaps there is something I am not aware of in YaST2, but it seems there are many settings which I need to edit by hand rather than through the rc.config editor. For example I don't know of an Apache configuration module. IIRC, there used to be some settings for Apache in the rc.config editor which are not available in any of the YaST2 configuration modules. It was often handy to be able to search the entire rc.config to find a setting. Sure I can grep (-r) for things in the /etc/sysconfig. I suspect newbies aren't going to be quite as eager to do things of this nature. I find icons and other graphical representations of configuration settings to be very helpful in figuring out how to set things up, and perhaps more importantly, to remember what I did. OTOH, I never want to be in a situation where I fill in a field click ok and 'magic happens'. I want to know what I changed, and where to find the setting in the associated config file. I also apprecieate comments in the config file which explain the particular setting, and even the perferd method of modifying it. I know that is a lot of up front work, especially when the native language of the developer is not English.
3. YaST2 sysconfig editor (like YaST2 rc.config editor before). Yes, I've been using that. One thing I found a bit confusing was the removal of many of the start variables for different services. You have the same three choices as before.
Perhaps you are correct. I just 'feel' like I don't.
Now it seems necessary to start processes such as Apache by explicitly modifying the runlevels. That's fine, if that's what SuSE intend for us to do. The interface we have for that is likewise very nice. The difficult part seems to be figuring out what SuSE expect us to do. That is, *should* I be using the runlevel editor for starting services automatically?
You can do what you wish. If you like to do it with the mouse, use the YaST2 runlevel editor. If you don't like this, use insserv or chkconfig on the console. Or use ln and rm, ...
Yes, I've done all of these at one time or another. I guess this goes back to 'what is the *SuSE* way?' Perhaps your answere is: 'whatever works for you'. I do like to have the freedom to do things in alternative ways, OTOH, there have been times when I undermined SuSEConfig, or got my changes shot out from under me by automated configuration mechanisms such as DHCP and/or ADSL changing my resolv.conf (that has been fixed, BTW). There was a time when some files were MD5 hashed to detect changes made by mechanisms other than SuSEconfig. That led to confusion when I made changes in both ways and my most recent YaST-based changes were not in effect.
Is there a paradigm shift in YaST? Am I just having trouble adjusting to a new and better way of doing things, or are there really shortcomings in the current (SuSE 8).
We only make some things easier: why creating/deleting links and edit START_* variables, if creating/deleting links is enough and more flexible? Now you can start services only for one runlevel and not, as before, for all or none.
This is great! I like the change, I just wasn't sure of how to use it.
I very much like the new look and feel of the GUI based YaST2. The terminal based version, well....(how 'bout using tab to move the focus?) What do others think of YaST verses YaST2?
I always use tab to move to the next focus.
Sorry, that wasn't what I should have said. I know there are some unintuitive aspects of that UI. I guess I do things without thinking about how I'm doing them. I think what used to bother me was that I didn't know to hold down the ALT key to select an option. That used to bug me when I saw the highlighted letters but they didn't seem to do anything. The thing I still find a bit awkward is the fact that when I have a seleciton highlighted, I still need to ALT-L or tab to [Launch]. Sure I know how to do these things, I just end up tabbing past my selection. NOW I REMEMBER! Please tell me the SHIFT+TAB feature wasn't there until recently. I'm almost certain that didn't used to work.
Thorsten
Steven
participants (3)
-
Leendert Meyer -
Steven T. Hatton -
Thorsten Kukuk